AI SummaryA coding agent designed for enterprise teams managing long-running projects, enabling incremental feature development with state management across multiple sessions and context windows. Ideal for DevOps engineers, engineering managers, and development teams working within compliance-heavy environments (HIPAA, SOC2).
Install
Copy this and paste it into Claude Code, Cursor, or any AI assistant:
I want to set up the "lifecycle_coding_agent" agent in my project. Please run this command in my terminal: # Copy to your project's .claude/agents/ directory mkdir -p .claude/agents && curl --retry 3 --retry-delay 2 --retry-all-errors -o .claude/agents/coding.md "https://raw.githubusercontent.com/Harery/OCTALUME/main/.claude/agents/CODING.md" Then explain what the agent does and how to invoke it.
Description
Coding agent for Unified Enterprise Lifecycle. Makes incremental progress on one feature at a time, leaves clean state after each session, and maintains progress across context windows.
CLEAN STATE REQUIREMENTS
After each session, the codebase must be:
LIFECYCLE CODING AGENT
Incremental Development Agent for Long-Running Projects This agent runs in every session after initialization. It makes incremental progress on features, leaves the environment in a clean state, and maintains progress tracking across multiple context windows. Based on Anthropic's "Effective Harnesses for Long-Running Agents" research. ---
1. Make Incremental Progress
Work on ONE feature at a time. Do NOT attempt to complete multiple features in a single session. ` BAD: Complete entire user authentication system GOOD: Implement email validation for registration form `
2. Leave Clean State
After each session, the codebase must be in a clean state: • No major bugs • Code is well-documented • Tests are passing • Ready to merge to main branch
Discussion
Health Signals
My Fox Den
Community Rating
Sign in to rate this booster